Low Impedance Pickup Research

[alexoest]

Hi Joe

I bent the wire into rough shape, filed the two joining areas flat for a larger contact area. Then I slid the CT in place and soldered the loop with a gas blowtorch using ordinary low temperature tin. I placed some damp paper tissue near the CT to absorb heat if it'd travel that far. Having three windings made it easier to get the CT far away from the soldering area than if I'd had only one winding.

Short circuits are avoided by painting the wire with nail polish (some my wife had left over from the early nineties. I later removed the nail polish except for where the turns are close to one another.

/Alex

[alexoest]

Hi Joseph

I will try that option with two CTs at some point. But I'm still to buy the microphone matching transformer.

I've been postponing it, expecting that I'd get my Linux audio computer up and running and record from the mic in on the sound card instead of using a guitar amp, but that hasn't happened yet.

/Alex

Alex,

Try one sting loop and a CT at each end of the loop. Then, try listening to the CTs in series and parallel and then again into a mic matching transformer. You should hear some interesting tonal variations doing this experiment.

It is good to see some daring people try this CT pickup stuff.

Joseph Rogowski

[alexoest]

Nice. The aluminium pickup is going to look really good when it's finished (I assume you'll remove the surplus thead and round the edges a bit)

How do you like the sound? (and do you use a microphone matching transformer?).

Plus, have you taken precautions against the oxidation of the contact points of the aluminium pickup, leading to greater resistance in the circuit (it was mentioned earlier in this thread)?

Btw. +1 on thanking bbsailor for sharing the idea. It's great fun to work with.

/Alex

Joseph, I probably wouldn't have tried this if you hadn't shared your work, so thank you

I made a couple for bass, both using a single AS-104 current transformer.

I used 1.6mm copper wire on this first one to check that it worked ok.

[ATTACH=CONFIG]12524[/ATTACH]

The second one was based on bajaman's design. I didn't have 5mm wide aluminium bars for the edge pieces, so had to settle for 10mm. I haven't completed this one yet and just held it over the strings to test, but it works OK. Still need to screw in the base plate I cut from some galvanised iron , which is why the magnets aren't sitting right yet, then cut the screws to size. The base plate is a great idea and does increase output slightly as you mentioned. Anyway, I'm going to use a second CT on this one to utilise the switching options as I do on my standard bass humbuckers.

[ATTACH=CONFIG]12525[/ATTACH]

This second effort was quite a bit of work.

BTW, I wired them into an Alembic Stratoblaster I have in my test bass to test them, didn't bother to check how they matched up.

[mkat]

Yep, will cut the screws once I account for the base plate and round the edges.

The tone is ok and like bajaman stated in the other thread on this, positioning is the main thing. I haven't tested this properly, I just held it over the strings while playing the open strings. Also, I'm running this through an Alembic Stratoblaster as stated above instead of a microphone matching transformer, as I don't have an one but do have a Stratoblaster installed in my test bass. I need to work out whether these things match up ok.

I'm going to route a cavity for this when I get a chance and do some proper testing. I haven't yet done anything to protect against oxidisation, as there is still some work to do with the aluminium (flush sand ends, round edges).

[Joe Gwinn]

I bent the wire into rough shape, filed the two joining areas flat for a larger contact area. Then I slid the CT in place and soldered the loop with a gas blowtorch using ordinary low temperature tin. I placed some damp paper tissue near the CT to absorb heat if it'd travel that far. Having three windings made it easier to get the CT far away from the soldering area than if I'd had only one winding.

Short circuits are avoided by painting the wire with nail polish (some my wife had left over from the early nineties. I later removed the nail polish except for where the turns are close to one another.

Ahh. That works, and explains the appearance in the photo.

[bbsailor]

Hi Joseph

I will try that option with two CTs at some point. But I'm still to buy the microphone matching transformer.

I've been postponing it, expecting that I'd get my Linux audio computer up and running and record from the mic in on the sound card instead of using a guitar amp, but that hasn't happened yet.

/Alex

Alex,

The microphone matching transformer is designed to do two things.
1. Boost the voltage output about 10 times.
2. Provide a better impedance match to the high impedance input of a guitar/bass amplifier.

If you are going directly into a microphone level input between about 1000 to 2000 ohms, then it should work just fine without the transformer.

For the lowest noise be sure to ground the low impedance string loop. I like using the Triad CSE-187L Current Transformer (CT) because I can make a tight low impedance string loop with a CT on each end. Plus, I can run a ground wire that is directly soldered to the CT metal frame as well a the low impedance string loop. Two of the CSE-187L CTs with parallel outputs match the input impedance of a microphone matching transformer or an XLR microphone input on a mixer pretty well. The low noise of this design allows the extra gain of the microphone mixer input circuit to boost the output to a good working level.


Joseph Rogowski

[Man Of Steel]

I'd like to thank and congratulate bbsailor for starting this thread, as well as all who contributed to it. I've been lurking for a couple of weeks, taking this in as time permits. For now, I'm giving the thread a bump while I read over the content thus far before I have anything else to offer.

[bbsailor]

Man Of Steel and all others interested in low impedance pickups,

Here is an update about some new findings retaled to current transformer low impdance pickups.
Using two current transformers, one CT is connected to the audio input and the other is used to tune the low impdance loop.

You can tune the low impedance loop by using two CSE-187L or similar current transformers (CT) with the low resistance leads facing each other and connected by 2.5 inches of AWG 12 solid wire, with a magnet in the center. Measure the output impdance of a single CT with a wire loop the same size and length as the wire in the CT primary. Then measure the impedance of the output CT with another CT forming the other side of the loop. It will be about 10 times higher.

WHY?
The theory goes that with a 500 to 1 turns ratio, the impedance ratio will be the square of the turn ratio or 250,000. Any slight change in the primary string loop impedance will be reflected back. A simple test is to use an alligator clip and short out the second CT and listen to the incresed low end response. You can put a 1K pot across the second CT and vary the tone of the loop because the variable rsistance is reflected back onto the primary loop loop impedance which is reflected again back into the output CT. The key is to listen to where the range of the pot value (across the second CT) makes pleasant tonal changes.

I hope this gives those who tinker with this stuff, something more to play and have fun with.



Joseph Rogowski

[Man Of Steel]

That's yet another neat idea, bbsailor -- thank you for passing that along, too!

It raises a couple of questions, like most great ideas:

1. Is the secondary of the second transformer, which I'll refer to as the tuning transformer, connected to anything, or is it allowed to float?

2. Could one achieve similar results replacing the tuning transformer with a ferrite slug and a wind or so of the 12 gauge wire, so that the necessary number of joints can be cut in half? I could imagine tuning this by sliding the core in or out.

Thanks!

[bbsailor]

That's yet another neat idea, bbsailor -- thank you for passing that along, too!

It raises a couple of questions, like most great ideas:

1. Is the secondary of the second transformer, which I'll refer to as the tuning transformer, connected to anything, or is it allowed to float?

2. Could one achieve similar results replacing the tuning transformer with a ferrite slug and a wind or so of the 12 gauge wire, so that the necessary number of joints can be cut in half? I could imagine tuning this by sliding the core in or out.

Thanks!

1. The tuning transformer raises the low impedance string loop. Putting a pot across the tuning transformer does the same thing as changing the size (diameter) of the low impedance string loop. Shorting out the tuning transformer secondary with a variable pot allows audible tuning of the pickup.

2. Ferrite slug is not variable. However, the tuning transformer offers a chance to alter low frequency response using the low impedance loop while using conventional capacitor values across the audio CT output to tame the high frequencies.

Also, you can wire an on-on-on type mini switch to select series, parallel or single CT output. In the single CT position the variable pot can be used to tune the pickup using the tuning transformer.

I hope this helps.

Joseph Rogowski

[Man Of Steel]

Thank you for your reply, Joseph!

1. The tuning transformer raises the low impedance string loop. Putting a pot across the tuning transformer does the same thing as changing the size (diameter) of the low impedance string loop. Shorting out the tuning transformer secondary with a variable pot allows audible tuning of the pickup.

If I understand correctly, this allows adjustment of the loading of the string loop. Aside from the elements of the tuning transformer's equivalent circuit, a 0-500 Meg pot functions, in essence, as a 0-2 ohm load on the string loop because of the 500:1 turns ratio. Am I understanding this correctly?

2. Ferrite slug is not variable. However, the tuning transformer offers a chance to alter low frequency response using the low impedance loop while using conventional capacitor values across the audio CT output to tame the high frequencies.

I realize it's not RF, but would not moving a permeable core closer to the center of the loop at the end of a single-transformer current-mode pickup increase the inductance of the pickup, effectively placing inductance in series? Many old Philco radios used ferrite slug tuning for presets.

"Taming the high frequencies" may be a fundamental issue with the low impedance paradigm. The transformer makes the DC resistance and inductance of the single turn transducer larger by the square of the turns ratio, but the stray capacitance is diminished by the same factor, which may be negligible to begin with. I suspect much of the capacitance comes from the transformer's secondary. The resonant frequency may be too high, and the Q-factor too low, to sound "right." Minimizing the equivalent series resistance of transducer loop and having enough shunt capacitance seem two (of several) key considerations. Fortunately, the latter is easy to fix.

Also, you can wire an on-on-on type mini switch to select series, parallel or single CT output. In the single CT position the variable pot can be used to tune the pickup using the tuning transformer.

More intriguing possibilities to consider -- thank you!

I hope this helps.

Joseph Rogowski

I'm still wrestling with an equivalent circuit of this beast with one transformer. We know that electrical chracteristics of a "conventional" voltage-mode pickup can be modeled as a second-order low-pass filter. For the current-mode pickup, the transformer adds an additional inductor, yielding a third-order filter. This raises the issue of alignment, if this inductance is significant (as well as the shunt stray capacitance). A random alignment is unlikely to perform well (i.e., sound good). Alignments for the "conventional" voltage-mode pickups evolved over the years in the form of winding specifications, magnet/polepiece sizes and shapes, and other factors. I imagine a lot of experimentation, trial, error, and tweaking went into the evolution of these alignments. This paradigm is so far out of my comfort zone (perhaps why it's so interesting?) that much of the conventional wisdom doesn't apply. I'm interested in modeling this current-mode type, plugging in design constraints, and playing around with free parameters to try to align them with less experimentation. Hopefully, come up with alignments that will allow me to actually put one I'll build on a guitar.

Thank you again!

[bbsailor]

Thank you for your reply, Joseph!



If I understand correctly, this allows adjustment of the loading of the string loop. Aside from the elements of the tuning transformer's equivalent circuit, a 0-500 Meg pot functions, in essence, as a 0-2 ohm load on the string loop because of the 500:1 turns ratio. Am I understanding this correctly?

Man Of Steel

Generally you should use a pot value that is betweeen 25 to 50 times the DC resistance of the pickup coil. But since the CSE187L current transformer has a max DCR of 21 ohms and has a closly coupled laminated core, the secondary impedance will be relativly higher than a traditionally wound guitar pickup of the same resistance. I would use a pot value that is between 1K to 2.5K ohms in value and listen to the "sweet spot" value where the change in tone in maximized within the pot range. The minimum DCR of the low impedance string loop with just two CSE187L CTs is 2 times 250 micro ohms plus the resistance of two straight pieces about 2" long to fit the spread of six strings. This puts the total resistance of the string loop to be below 1000 micro ohms or 1 milliohm. Check out Surplus Sales of Nebraska for some rectangular copper wire that is 0.162" on a side and can be drilled to accomodate the "U-Shaped" primary lead with a snug fit to ensure a good, low resistance connection. The lower the total string loop resistance the more efficient the tuning CT will be in changing to tone.

I hope this points you in the right direction.

Joseph Rogowski

[Man Of Steel]

Progress: Modeling a low-current pickup

Thank you, Joseph.

An update on modeling: I believe the current-mode pickup may be modeled as a cascade of two-port subnetworks. The model will be slightly simpler if the inductance and series resistance of the transformer primary can be combined with the corresponding quantities of the secondary. It will also be slightly simpler if there are no eddy losses in the single "under string" winding. If both simplifications apply, it will be a fifth-order model. Otherwise, bump up the order of the model by one for each of these assumptions that are not applied.

This is, of course, before the tone control, cable, etc. are considered.

I think I can provide an Octave m-file (possibly compatible with Octave's proprietary counterpart, Matlab) that can compute the electrical transfer function as a function of frequency for a given set of parameters. These parameters are:

1. Under-string inductance;
2. Under-string resistance;
3. Under-string capacitance;
4. Shunt capacitance, series resistance, and series inductance of transformer primary;
5. Shunt capacitance, series resistance, and series inductance of transformer secondary;
6. Transformer coupling factor;
7. Transformer turns ratio;
8. Volume control resistance and setting;
9. Tone control capacitance, resistance, and setting;
10. Load resistance;
    and, if necessary:
11. Leakage resistance and effective fraction of coil bypassed by eddy current;
12. Load (cable) capacitance

Many of the parameters relate to the transformer. The Triad data sheet for the CSE187L transformer provides information useful for its intended application of current-to-voltage conversion, but, other than the turns ratio and DC resistances, does not help here. (The fact that the core starts saturating at 10 to 20 amperes, depending on frequency, is reassuring, but is hardly an issue with a guitar pickup!) Has anybody evaluated the primary and secondary capacitances, primary and secondary inductances, and equivalent core loss resistance of this transformer? It would be a great help if they could share such information with me!

[bbsailor]

Thank you, Joseph.

An update on modeling: I believe the current-mode pickup may be modeled as a cascade of two-port subnetworks. The model will be slightly simpler if the inductance and series resistance of the transformer primary can be combined with the corresponding quantities of the secondary. It will also be slightly simpler if there are no eddy losses in the single "under string" winding. If both simplifications apply, it will be a fifth-order model. Otherwise, bump up the order of the model by one for each of these assumptions that are not applied.

This is, of course, before the tone control, cable, etc. are considered.

I think I can provide an Octave m-file (possibly compatible with Octave's proprietary counterpart, Matlab) that can compute the electrical transfer function as a function of frequency for a given set of parameters. These parameters are:

1. Under-string inductance;
2. Under-string resistance;
3. Under-string capacitance;
4. Shunt capacitance, series resistance, and series inductance of transformer primary;
5. Shunt capacitance, series resistance, and series inductance of transformer secondary;
6. Transformer coupling factor;
7. Transformer turns ratio;
8. Volume control resistance and setting;
9. Tone control capacitance, resistance, and setting;
10. Load resistance;
    and, if necessary:
11. Leakage resistance and effective fraction of coil bypassed by eddy current;
12. Load (cable) capacitance

Many of the parameters relate to the transformer. The Triad data sheet for the CSE187L transformer provides information useful for its intended application of current-to-voltage conversion, but, other than the turns ratio and DC resistances, does not help here. (The fact that the core starts saturating at 10 to 20 amperes, depending on frequency, is reassuring, but is hardly an issue with a guitar pickup!) Has anybody evaluated the primary and secondary capacitances, primary and secondary inductances, and equivalent core loss resistance of this transformer? It would be a great help if they could share such information with me!

Man of Steel,

You are on the right track with your anaysis, but there are some more details that will help you form a good mental model and ultimately a good math model that will define and graph what is actually happening.

Remember that a transformer reflects the secondary load back to the primary by the square of the turns ratio (TR). So, in the CSE187L with a 1:500 TR you are reflecting a secondary load impedance back in parallel with the primary string loop impedance which is already pretty low in the sub milliohm range typically measured in hundreds of a microohm. This cannot be directly measured but what you can do is to measure the current transformer (CT) output AC resistance with the Extech LCR meter with the CT having (1) an open primary and (2) having a low imepdance string loop connected. Now, add a second CT on the other end of the same length string loop and note that the second CT will increase the output impedance of the output CT connected to the Extech LCR meter. Assume that the CT output is being sent into a mic input impedance of 2K ohms. Then, a reflected impedance of 2K divided by 250,000 (500 squared) or 0.008 ohms (8 milliohms or 8,000 microohms) will appear in parallel with the string loop impedance. Changing the gauge of the primary string loop will help tune the impedance range of the output CT. If you are going into an XLR input or a remotely mounted microphone matching transformer the microphone matching transformer input load can be optimized so that when the matching tranformer load impedance is reflecte back into the primary string loop, various harmonic ranges can be optimized to satisfy what the ear considers as being pleasing.

Lets take a more extreme case using a Prem Magnetics SPCT-251 CT with a TR of 1:2000 makes a 4,000,000 string reflected impedance. If you use two 5" lengths of .162 square copper wire, which easily fits into the SPCT-251 open primary space. Drill two .12" holes on each end of the 5" lengths to accomodate a solid piece of copper wire with a CT on each end forming a few hundred microohm primary loop. Look at the output AC resistance of once CT with the other CT open and you will see about a 20K ohm AC resistance both at 1000Hz and almost the same at 120Hz. Now, short out the second CT with a short jumper and see the reflected AC resistance will be about 10X lower or about 2K ohms on the Extech LCR meter. This is due to primary loop impedance being changed by the second CT reflecting a short back into the primary loop by an impedance ratio of 4,000,000, minus the leakage inductance but this shows how the gauge of the wire on the primary loop can change the tonal characteristics of the two-CT-string-loop under the guitar strings.

Velleman makes a 2-channel scope with a function generator PCSGU250 and software package PC-Lab 2000LT that can run Bode plots by feeding the signal generator into a 500 turn coil of AWG 32 wire on a HB pickup bobbin to stimulate the CT assembly. Try it with a CT open and then again with the CT shorted. Switch the outputs of the two CTs in series and parallel with a simple DPDT switch and listen to the tonal changes. Use an on-on-on swith to make a series/parallel/single coil selection. In the single CT mode, put a variety of pot loads across the second CT as wall as a variety of capacitors and listen to the tonal differences and then note where the most pleasing range of tone changes occurs and then run the bode plot to actually see what is happening. This will be most educational as well as open you mind to what is actually happening in an extremely low impedance string loops with a variety of CTs with different turns ratios. The SPCT-251 will provide you with about 100mv peak output from the 305 ohm secondary coil. Only the ear will tell you what sounds good, then you can graph and print out the Bode plots to capture a picture of the frequency response of what sounds pleasing and then see why it is pleasing.

Guitar strings tend to emphasize certain harmonics but since typical high impedance pickups with 6 to 10 thousand turns of very thin wire (AWG 42, 43) tend to be very low Q coils and have a low inductance. Enter the low impedance CT pickup domain and you can see new opportunities to voice the sound of the pickup by selecting the right gauge of the low impedance string loop as well as tuning that low impedance string loop with the second CT helping to tune the primary loop impedance. This analysis uses traditional transformer theory but because CTs have very high turns ratios, reflected loads can have a very noticable effect on the final pickup sound and the harmonics that are emphasized by the combination of CTs types, turns ratios, primary loop wire size, second CT loading and personal tonal preferences. Ultimately, you will notice that the extended frequency response can be well out beyond the 5KHz to 8KHz audio region where high impedance pickups do not produce very much upper harmonic content.


Lace, the makers of the Alumitone pickup, is missing an opportunity to put a second set of coils on the opposite side of the Alumitone frame where the second CT can be modeled to tune the pickups to sound new ways without needing active electronice. Lace can also put the two sets of secondary outputs in series, parallel or single (as explained above). I estimate that the tiny coils under the frame of the Alumitone frame have about 10,000 turns of wire smaller in diamter than typical AWG 42 or 43. Hint, if Don Lace is reading this post, this is a free design idea to expand the Alumitone pickup design to fully capture it's full design potential. This design concept is all based on using the second CT to tune the low impedance string loop formed by the Alumitone alumium frame.


This is just some more "food for thought" for those curious enough to play with CT pickup designs. Also, this design can make very good sounding magnetic acoustic guitar pickups without the classic resonant hump in the mid audio range that cuts off the upper harmonics that usually helps to define an acoustic guitar sound not adequtely sensed by high impedance magnetic passive pickups.

Joseph Rogowski

[MoreBass]

The Lace Alimatone pickups caught my interest recently when I was researching some info related to pickup replacement on my Bass guitar. After reading some of these threads on constructing a Lo-Z pickup (thanks to contributing members, especially bbsailor) and being an occasional DIYer, I ordered a couple of CSE187L transformers and constructed the pickup pictured below. I physically designed the layout around a magnet I scrounged from a discarded magneto-optical drive (guessing neodymium) looped through 2 Low Z transformers with 8 awg copper wire. Output was fed to a line matching transformer with an XLR cord. Volume wise I thought the output was about the same as my passive pickups, but the tone sounded hollow and very trebly. Any suggestions on how to improve construction to enhance low end?